Libraries
pacman::p_load(haven,
ggplot2,
tidyverse,
viridis,
plotly)
Data
setwd("~/Documents/GitHub/Final_Data_Science_Project/News source df's")
dat <- read_csv("data_set_full.csv") %>%
select(-INCOME_GRP, -count) %>%
mutate(rate = as.numeric(rate)) %>%
rename(Country = ADMIN2)
dat_top_rates <- dat %>%
group_by(Country) %>%
summarise(tot_rate = sum(rate)) %>%
ungroup()
dat <- left_join(dat, dat_top_rates) %>%
arrange(desc(tot_rate))
Most mentioned countries by center-right newspapers
dat_right <- dat %>%
filter(orientation == "Center right", Country != "Oman") %>%
head(67) %>%
mutate(Country = fct_reorder(Country, tot_rate))
dat_right$Country <- factor(dat_right$Country,
levels = c("Hong Kong S.A.R.", "Afghanistan", "France", "Germany", "South Africa", "Democratic Republic of the Congo", "Japan", "Italy", "United States of America", "Russia", "China", "Iran", "Australia", "India", "United Kingdom"))
personal_theme = theme(plot.title =
element_text(hjust = 0.5))
plot_right <- dat_right %>%
ggplot(aes(x = Country, y = rate, fill = source)) +
geom_col(color = "black", size = 0.1, width = .5) +
ggtitle("15 most mentioned countries by center-right leaning newspapers") +
theme_minimal() +
labs(y ="Proportion of country mentions on total headlines (%)", fill = "Source") +
scale_fill_viridis(option = "rocket", discrete = T) +
coord_flip() +
personal_theme
ggplotly(plot_right)
Most mentioned countries by center-left newspapers
dat_left <- dat %>%
filter(orientation == "Center left", Country != "Oman") %>%
head(90)
dat_left$Country <- factor(dat_left$Country,
levels = c("South Africa", "France", "Germany", "Democratic Republic of the Congo", "Japan", "Australia", "India", "Russia", "Afghanistan", "Iran", "Italy", "United States of America", "China", "Hong Kong S.A.R.", "United Kingdom"))
plot_left <- dat_left %>%
ggplot(aes(x = Country, y = rate, fill = source)) +
geom_col(color = "black", size = 0.1,width = .5) +
ggtitle("15 most mentioned countries by center-left leaning newspapers") +
theme_minimal() +
labs(y ="Proportion of country mentions on total headlines (%)", fill = "Source") +
scale_fill_viridis(discrete = T) +
coord_flip() + personal_theme
ggplotly(plot_left)